Análisis de National Oil Company Database

UNGS/CONICET

Link National Oil Company Database

library(readr)
library(tidyverse)
library(ggplot2)
library(plotly)
library(lubridate)
options (scipen = 999)
  
#cargo la base
nocs_full <- read_csv( "../data/nocs/NRGI-NOCdatabase-FullDataset.csv") %>% 
   mutate(observation = as.double(observation),
          anio = (as.Date(parse_date_time(year, orders  = "%Y"))))

hidrocarburos_df <- read_csv("../data/balances/hidrocarburos_df.csv") %>% 
  select(-"...1")  
 
  
  

Variables que contiene la base

variables = unique(nocs_full$indicatorName)
variables

 [1] "Any shares publicly traded?"                                                      
 [2] "Auditor opinion issued with qualification?"                                       
 [3] "Bonus payments"                                                                   
 [4] "Capex (company-wide) per barrel"                                                  
 [5] "Capex / total revenue"                                                            
 [6] "Capital expenditures"                                                             
 [7] "Cash and cash equivalents"                                                        
 [8] "Cash flows from financing activities"                                             
 [9] "Cash flows from investing activities"                                             
[10] "Cash flows from operating activities"                                             
[11] "Cash ratio"                                                                       
[12] "Country released EITI report for this year?"                                      
[13] "Current assets"                                                                   
[14] "Current liabilities"                                                              
[15] "Dividends"                                                                        
[16] "Domestic oil, gas & product sales"                                                
[17] "Domestic sales"                                                                   
[18] "Employees"                                                                        
[19] "Equity"                                                                           
[20] "Exchange rate"                                                                    
[21] "Exploration - Seismic (2D)"                                                       
[22] "Exploration - Seismic (3D and 4D)"                                                
[23] "External oil, gas & product sales"                                                
[24] "External sales"                                                                   
[25] "Fiscal payments from contractors"                                                 
[26] "Gas production"                                                                   
[27] "Gas production of home country"                                                   
[28] "GDP"                                                                              
[29] "General government revenue"                                                       
[30] "General government total expenditure"                                             
[31] "Government gross debt"                                                            
[32] "Government resource revenue / general government revenue"                         
[33] "Government transfers"                                                             
[34] "Income tax"                                                                       
[35] "Longterm/fixed assets"                                                            
[36] "National gas reserves"                                                            
[37] "National oil and gas reserves"                                                    
[38] "National oil reserves"                                                            
[39] "Natural capital, subsoil assets: oil & gas"                                       
[40] "Net income after taxes"                                                           
[41] "Net income after taxes/equity"                                                    
[42] "Net income after taxes/revenue"                                                   
[43] "Net income after taxes / total assets"                                            
[44] "Net income from all revenues (before transfers to govt.)"                         
[45] "Net income from core revenues (before transfers to govt.)"                        
[46] "Net income per employee"                                                          
[47] "Net income / total revenue (before transfers to govt.)"                           
[48] "NOC debt / government gross debt"                                                 
[49] "NOC has stake in fields abroad?"                                                  
[50] "NOC net income / general government revenue"                                      
[51] "NOC oil and gas production / oil and gas production of home country"              
[52] "NOC operator of any production?"                                                  
[53] "NOC operator of fields abroad?"                                                   
[54] "NOC reserves / national oil and gas reserves"                                     
[55] "NOC total assets / total national wealth"                                         
[56] "NOC total revenues / GDP"                                                         
[57] "NOC total revenues / general government revenues"                                 
[58] "NOC transfers to government / general government revenue"                         
[59] "NOC transfers to government / NOC net income"                                     
[60] "NOC transfers to government / total NOC revenues"                                 
[61] "NOC transfers to government / total public expenditures"                          
[62] "O&G production / total reserves"                                                  
[63] "O&G revenue per barrel of production"                                             
[64] "O&G revenue per barrel of reserves"                                               
[65] "O&G revenue per employee"                                                         
[66] "Oil and gas production of home country"                                           
[67] "Oil & gas production"                                                             
[68] "Oil, gas & product sales"                                                         
[69] "Oil production"                                                                   
[70] "Oil production of home country"                                                   
[71] "Operational expenditures"                                                         
[72] "Opex (company-wide) per barrel"                                                   
[73] "Opex / total revenue"                                                             
[74] "Other transfers to government"                                                    
[75] "Percentage non-core activities in total revenue"                                  
[76] "Proceeds of state profit/equity petroleum"                                        
[77] "Production on which NOC is operator"                                              
[78] "Production per employee"                                                          
[79] "Report audited by independent external auditor?"                                  
[80] "Report presented according to International Financial Reporting Standards (IFRS)?"
[81] "Reserves"                                                                         
[82] "Reserves per employee"                                                            
[83] "Reserves / production ratio"                                                      
[84] "Return on capital employed (before transfers to govt.)"                           
[85] "Revenue - Non core activities"                                                    
[86] "Rigs"                                                                             
[87] "Royalties"                                                                        
[88] "Total assets"                                                                     
[89] "Total liabilities"                                                                
[90] "Total longterm/fixed liabilities"                                                 
[91] "Total national wealth"                                                            
[92] "Total revenue"                                                                    
[93] "Total revenue per barrel of production"                                           
[94] "Total revenue per employee"                                                       
[95] "Total transfers to government"                                                    
[96] "Transfers to government per barrel"                                               
[97] "Wells drilled"                                                                    

Lista de todas las empresas

#empresas disponibles
unique(nocs_full$company)

Cantidad de empresas

length(unique(nocs_full$company))

#filtro exploratorio de empresas 
filtro_empresas <- c("YPF", "Petrobras", "YPFB", "PDVSA", "Saudi Aramco")

#filtro con empresas de América Latina
empresas_AL <- nocs_full %>% 
  filter(region ==  "Latin America/Caribbean") %>% 
  select(company, country) 
empresas_AL <- empresas_AL[!duplicated(empresas_AL), ]

empresas_AL

NA

Producción

ggplotly(production_graf)

The shape palette can deal with a maximum of 6 discrete values because more than 6 becomes difficult to discriminate; you have
61. Consider specifying shapes manually if you must have them.

Activo y Patrimonio Neto

assets_graf <-  nocs_full %>% 
  filter(indicatorName %in% c( "Total assets"),
         year == 2017, units == "USD million",
         !is.na(observation)) %>%
  ggplot(aes(x = reorder(company,-observation) , y= observation, fill = company))+
  geom_col(position = "stack")+
  labs(title =  "Total assets",
       x = "Empresa", y = "MM USD")+
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 90, hjust = 1))
  # facet_grid(indicatorName~region, labeller = label_wrap_gen(width=17))
ggplotly(assets_graf)

NA

assets_graf <-  nocs_full %>% 
  filter(indicatorName %in% c( "Equity"),
         year == 2017, units == "USD million",
         !is.na(observation)) %>%
  ggplot(aes(x = reorder(company,-observation) , y= observation, fill = company))+
  geom_col(position = "stack")+
  labs(title =  "Total Equity", subtitle= "Year 2017",
       x = "Empresa", y = "MM USD")+
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 90, hjust = 1))
  # facet_grid(indicatorName~region, labeller = label_wrap_gen(width=17))
ggplotly(assets_graf)

Inversión en capital

capex_graf <-  nocs_full %>% 
  filter(indicatorName %in% c("Capital expenditures"), units == "USD million") %>%
  ggplot(aes(x = year, y= observation, color = company))+
  geom_line(aes(linetype = country))+
  labs(title =  "Capital expenditures (CAPEX)",
       x = "Año", y = "MM USD")+
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 90, hjust = 1),
        strip.text.y = element_text(size = 7))+
  facet_grid(~region, labeller = label_wrap_gen(width=15))

ggplotly(capex_graf)

Rentabilidad

Data Frame con indicadores

#para cálculo de rentabilidad
nocs_df <- nocs_full %>%
  select(year, company, region, country, productionGroup, indicatorName, units, observation) %>% 
  filter(indicatorName %in% c("Longterm/fixed assets", "Cash and cash equivalents",
                               "Net income after taxes", "Income tax",
                              "Equity","Total assets", 
                              "Total liabilities" , "Current liabilities" ),
         units == "USD million") %>%
  mutate(row_id = 1:nrow(.)
         # , observation = as.double(observation)
         ) %>%
  spread(.,
         key = indicatorName,
         value = observation) %>% 
  select(-c(row_id)) %>% 
  rename( fixed_assets = "Longterm/fixed assets", cash_and_equi = "Cash and cash equivalents",
         net_income_after_tax = "Net income after taxes" , income_tax = "Income tax",
         equity = "Equity",  assets = "Total assets",
         liabilities = "Total liabilities", current_liabilities = "Current liabilities") %>% 
  mutate(net_income_after_tax = as.double(net_income_after_tax),
         fixed_assets = as.double(fixed_assets),
         cash_and_equi = as.double(cash_and_equi),
         income_tax = as.double(income_tax),
         equity = as.double(equity),
         liabilities = as.double(liabilities),
         year = parse_date_time(year, orders = "y")) %>% 
  group_by(year, company, country, region, productionGroup, units) %>%
  summarise(fixed_assets = sum(fixed_assets, na.rm = T),
            cash_and_equi = sum(cash_and_equi, na.rm = T),
            net_income_after_tax = sum(net_income_after_tax, na.rm = T),
            income_tax = sum(income_tax, na.rm =  T),
            assets = sum(assets, na.rm =  T),
            liabilities = sum(liabilities, na.rm =  T),
            current_liabilities = sum(current_liabilities, na.rm =  T),
            equity = sum(equity, na.rm =  T),
            KTA = cash_and_equi + fixed_assets,
            tg_beforetax = (income_tax + net_income_after_tax) / KTA,
            tg_aftertax = net_income_after_tax / KTA,
            ratio_endeudamiento = liabilities / equity, 
            ratio_endeudamiento_cp = current_liabilities/equity,
            pasivo_activo = liabilities/assets,
            ratio_solvencia =  equity /liabilities) 

# nocs_df %>% arrange(-tg_aftertax) %>% 
#   select(company, KTA, tg_aftertax)

nocs_df

nocs_df[sapply(nocs_df, is.infinite)] <- NA

Rentabilidad media de la rama

tg_media <- nocs_df %>%
    group_by(year) %>% 
    summarise(
                  tg_media_before_tax = sum(income_tax, net_income_after_tax, na.rm =  T)/ 
                                                        sum(fixed_assets, cash_and_equi, na.rm = T),
                  tg_media_after_tax = sum(net_income_after_tax, na.rm =  T) / 
                                                        sum(fixed_assets, cash_and_equi, na.rm = T)) %>% 
                  # tg_media_before_tax_mean = sum(mean(income_tax, na.rm =  T), 
                  #                                mean(net_income_after_tax, na.rm = T), na.rm = T)/ 
                  #                             sum(mean(fixed_assets, na.rm = T), 
                  #                                 mean(cash_and_equi, na.rm = T), na.rm = T),
                  # tg_media_after_tax_mean = mean(net_income_after_tax, na.rm = T)/ 
                  #                             sum(mean(fixed_assets, na.rm = T), 
                  #                                 mean(cash_and_equi, na.rm = T), na.rm = T)) %>% 
     gather(.,
               key = variable,
               value = valor, 
               2:3) 
    # mutate(metodo = case_when( str_sub(variable[-3])))
                

tg_media_graf <- tg_media %>% 
  # filter(variable == "tg_media_before_tax_mean") %>% 
  ggplot(aes(year, valor, color = variable))+
  geom_line()+
  labs(title =  "NOC's. Tasa de Ganancia media")+
  theme(legend.position = "bottom")+
  scale_y_continuous(labels = scales::percent)
tg_media_graf

NA

nocs_df_con_media <- nocs_df %>% 
  bind_rows(tg_media %>% 
              spread(key = variable,
                     value = valor) %>% 
              mutate(company = "TG Media") %>% 
              rename(tg_beforetax = tg_media_before_tax,
                     tg_aftertax = tg_media_after_tax)) %>% 
  mutate(is_media = case_when(company == "TG Media" ~ "media",
                              TRUE ~ "empresas"))

write.csv(nocs_df_con_media, file = "nocs_df_con_media.csv")

graf_media_line <- nocs_df_con_media %>% 
  ggplot(aes(year, tg_aftertax, color = company, group = is_media))+
  geom_line()+
  geom_point()+
  labs(title =  "NOC's. Rentabilidad media y del resto de las empresas después de impuestos")+
  theme(legend.position = "none")+
  scale_y_continuous(labels = scales::percent, limits = c(-0.30, 0.70),breaks = seq(-0.30, 0.70, .1))
graf_media_line

graf_media_box <- nocs_df %>% 
  # ggplot(aes(year, tg_aftertax, color = company, size = is_media))+
  ggplot(aes(year, tg_aftertax, color = company, group = year))+
  geom_boxplot()+
  # ggplot(aes(data = nocs_df_con_media %>% filter(company == "TG Media"), year, tg_aftertax))+
  # geom_line()+
  labs(title =  "NOC's. Rentabilidad despues de impuestos")+
  theme(legend.position = "none")+
  scale_y_continuous(labels = scales::percent, limits = c(-0.3, .8), breaks = seq(-0.3, .8, 0.1))
graf_media_box

summary(nocs_df %>% 
  group_by(year, productionGroup) %>% 
  summarise(tg_aftertax = mean(tg_aftertax, na.rm = T)) %>% 
  spread(.,
         key = productionGroup,
         value =  tg_aftertax))

`summarise()` has grouped output by 'year'. You can override using the `.groups` argument.

      year                     Internationalized operators Large domestic producers Medium domestic producers Pre-production NOCs
 Min.   :2011-01-01 00:00:00   Min.   :0.0921              Min.   :0.1214           Min.   :0.03228           Min.   :-0.19649   
 1st Qu.:2012-10-01 12:00:00   1st Qu.:0.1124              1st Qu.:0.1677           1st Qu.:0.13562           1st Qu.:-0.03020   
 Median :2014-07-02 12:00:00   Median :0.1327              Median :0.2683           Median :1.35633           Median : 0.05986   
 Mean   :2014-07-02 12:00:00   Mean   :0.1424              Mean   :0.2681           Mean   :1.46370           Mean   : 0.20272   
 3rd Qu.:2016-04-01 12:00:00   3rd Qu.:0.1706              3rd Qu.:0.3280           3rd Qu.:2.72462           3rd Qu.: 0.13376   
 Max.   :2018-01-01 00:00:00   Max.   :0.2127              Max.   :0.4527           Max.   :3.17724           Max.   : 1.48343   
 Small domestic producers
 Min.   :-7.1290         
 1st Qu.: 0.0129         
 Median : 0.2032         
 Mean   :-0.6546         
 3rd Qu.: 0.3530         
 Max.   : 0.7303         

Rentabilidad según tamaño

tg_graf_prod_group <- nocs_df %>% 
  # filter(tg_aftertax != 0 & fixed_assets > 0 & tg_aftertax < 5 & tg_aftertax > -20 ) %>%  
  ggplot(aes(year, tg_aftertax, color = company))+
  geom_line()+
  labs(title =  "NOC's. Rentabilidad despues de impuestos. División por tamaño")+
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 90, hjust = 1))+
  scale_y_continuous(labels = scales::percent, limits = c(-0.3, .8), breaks = seq(-0.3, .8, 0.20))+
  facet_grid(~productionGroup,  labeller = label_wrap_gen(width=17))
plotly::ggplotly(tg_graf_prod_group)

Rentabilidad según región

tg_graf_xregion <- nocs_df %>% 
  filter(tg_aftertax != 0 & fixed_assets > 0 & tg_aftertax < 1 & tg_aftertax > -20 ) %>%
  ggplot(aes(year, tg_aftertax, color = company))+
  geom_line()+
  labs(title =  "NOC's. Rentabilidad despues de impuestos. División por regións")+
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 90, hjust = 1))+
  scale_y_continuous(labels = scales::percent)+
  facet_grid(~region,  labeller = label_wrap_gen(width=17))

plotly::ggplotly(tg_graf_xregion)

NA

YPF, Petrobras y empresas seleccionadas

ggplotly(tg_graf_ypf)

`geom_smooth()` using method = 'loess' and formula 'y ~ x'
Removed 8 rows containing non-finite values (stat_smooth).

# ruta_graf_tg <- "tg_filtro_graf.png"
# png(ruta_graf_tg)
# print(tg_filtro_graf)
# dev.off()

Analisis de correlación

#Correlación entre KTA y ganancias después de impuestos
cor(x = nocs_df$KTA, 
    y = nocs_df$net_income_after_tax, use = "complete.obs",
    method = "spearman")

[1] 0.7450791

cor(x = nocs_spread_2$capex, 
    y = nocs_spread_2$net_income, use = "complete.obs",
    method = "spearman")

[1] 0.6563327

Endeudamiento

Solvencia

\[ Solvencia = \frac{Patrimonio Neto}{Pasivo}\]

nocs_df %>% 
  ggplot(aes(year, ratio_solvencia ,group = year))+
  geom_boxplot() +
  # geom_point(aes(color = company))+
  # geom_line(aes(group = company))+
  geom_line(data = nocs_df %>% filter(company == "YPF"), 
            aes(year, ratio_solvencia,color = "red", group = company), size = 1)+
  theme(legend.position = "none")+
  labs(title = "Distribución de ratio de solvencia para todas las empresas e YPF")

ggplot(nocs_df %>% filter(company == "YPF"))+
  geom_line( aes(year, ratio_solvencia, group = company))+
  labs(title = "Ratio de solvencia de YPF")

solvencia_graf <- nocs_df %>% 
  filter(ratio_solvencia >=0 & ratio_solvencia < 4) %>% 
  ggplot(aes(year, ratio_solvencia, color = company))+
  geom_point(alpha = 0.5)+
  labs(title =  "Ratio de solvencia todas las empresas")+
  theme(legend.position = "none")+
  # scale_y_continuous(labels = scales::percent)+
  facet_wrap(~productionGroup)
ggplotly(solvencia_graf)

solvencia_graf_2 <- nocs_df %>% 
  # filter(ratio_solvencia >=0 & ratio_solvencia < 4) %>% 
  filter(company %in% empresas_AL$company) %>% 
  ggplot(aes(year, ratio_solvencia, color = company))+
  geom_point(alpha = 0.5)+
  labs(title =  "Ratio de solvencia de empresas de AL")+
  theme(legend.position = "none")+
  # scale_y_continuous(labels = scales::percent)+
  facet_wrap(~productionGroup)

ggplotly(solvencia_graf_2)

solvencia_filtro_graf <- filtro_emp_df %>% 
  # filter(ratio_solvencia >=0 & ratio_solvencia < 4) %>% 
  ggplot(aes(year, ratio_solvencia, color = company))+
  geom_line()+
  geom_point(alpha = 1, size = 3)+
  labs(title =  "Ratio de solvencia empresas seleccionadas")+
  theme(legend.position="bottom")
  # scale_y_continuous(labels = scales::percent)+

solvencia_filtro_graf

Endeudamiento

Ratio de endeudamiento

\[ Endeudamiento = \frac{Pasivo}{Patrimonio Neto} \]

nocs_df %>% 
  ggplot(aes(year, ratio_endeudamiento ,group = year))+
  geom_boxplot() +
  # geom_point(aes(color = company))+
  # geom_line(aes(group = company))+
  geom_line(data = nocs_df %>% filter(company == "YPF"), 
            aes(year, ratio_endeudamiento,color = "red", group = company), size = 1)+
  theme(legend.position = "none")+
  labs(title = "Distribución de ratio de endeudamiento de NOC's e YPF")+
  ylim(NA, 15 )

ggplot(nocs_df %>% filter(company == "YPF"))+
  geom_line( aes(year, ratio_endeudamiento, group = company))+
  labs(title = "Ratio de endeudamiento de YPF")

nocs_df %>% 
  ggplot(aes( ratio_endeudamiento ))+#,group = year))+
  geom_density() +
  labs(title = "Distribución de ratio de endeudamiento de NOC's")+
  facet_wrap(~year(year), scales = "free")

# tapply(nocs_df$ratio_endeudamiento, nocs_df$year, summary)

library(data.table)
nocs_dt = as.data.table(nocs_df)
nocs_dt[, as.list(summary(ratio_endeudamiento)), by =year]

nocs_df %>% 
  ggplot(aes(year, pasivo_activo ,group = year))+
  geom_boxplot() +
  # geom_point(aes(color = company))+
  # geom_line(aes(group = company))+
  geom_line(data = nocs_df %>% filter(company == "YPF"), 
            aes(year, pasivo_activo,color = "red", group = company), size = 1)+
  theme(legend.position = "none")+
  labs(title = "Distribución de Pasivo/Activo de NOC's e YPF")

  # ylim(NA, 15 )

ggplot(nocs_df %>% filter(company == "YPF"))+
  geom_line( aes(year, pasivo_activo, group = company))+
  labs(title = "Pasivo/Activo de YPF")

nocs_df %>% 
  ggplot(aes(year, ratio_endeudamiento_cp ,group = year))+
  geom_boxplot() +
  # geom_point(aes(color = company))+
  # geom_line(aes(group = company))+
  geom_line(data = nocs_df %>% filter(company == "YPF"), 
            aes(year, ratio_endeudamiento_cp,color = "red", group = company), size = 1)+
  theme(legend.position = "none")+
  labs(title = "Distribución de ratio de endeudamiento CP de NOC's e YPF")+
  ylim(NA, 12.5 )

ggplot(nocs_df %>% filter(company == "YPF"))+
  geom_line( aes(year, ratio_endeudamiento_cp, group = company))+
  labs(title = "Ratio de endeudamiento CP de YPF")

NA
NA

endeudamiento_graf <- nocs_df %>% 
  filter(ratio_endeudamiento <10 & ratio_endeudamiento > -3 ) %>% 
  ggplot(aes(year, ratio_endeudamiento, color = company))+
  geom_point(alpha = 0.5)+
  labs(title =  "Ratio de endeudamiento todas las empresas")+
  theme(legend.position = "none")+
  # scale_y_continuous(labels = scales::percent)+
  facet_wrap(~productionGroup)
ggplotly(endeudamiento_graf)

endeudamiento_graf_al <- nocs_df %>% 
  filter(ratio_endeudamiento <30 ) %>%
  filter(company %in% empresas_AL$company ) %>% 
  ggplot(aes(year, ratio_endeudamiento, color = company))+
  geom_point(alpha = 0.5)+
  labs(title =  "Ratio de endeudamiento empresas AL")+
  theme(legend.position = "none")+
  # scale_y_continuous(labels = scales::percent)+
  facet_wrap(~productionGroup)
ggplotly(endeudamiento_graf_al)

endeudamiento_filtro_graf <- filtro_emp_df %>% 
  # filter(ratio_solvencia >=0 & ratio_solvencia < 4) %>% 
  ggplot(aes(year, ratio_endeudamiento, color = company))+
  geom_point(alpha = 1, size = 3)+
  geom_line()+
  labs(title =  "Ratio de endeudamiento empresas seleccionadas")+
  theme(legend.position="bottom")
  # scale_y_continuous(labels = scales::percent)+

endeudamiento_filtro_graf

NA
NA

Valores absolutos

endeudamiento_all <- nocs_df %>% 
  # filter(company %in% empresas_AL$company ) %>%
  ggplot(aes(year, liabilities, color = company))+
  geom_point(alpha = 0.5)+
  geom_line()+
  labs(title =  "Liabilities", y = "USD million")+
  theme(legend.position="none")+
  facet_wrap(~productionGroup)
  # scale_y_continuous(labels = scales::percent)+

ggplotly(endeudamiento_all)

endeudamiento_al <- nocs_df %>% 
  filter(company %in% empresas_AL$company ) %>%
  ggplot(aes(year, liabilities, color = company))+
  geom_point(alpha = 0.5)+
  geom_line()+
  labs(title =  "Liabilities", y = "USD million")+
  theme(legend.position="none")+
  facet_wrap(~productionGroup)
  # scale_y_continuous(labels = scales::percent)+

ggplotly(endeudamiento_al)

endeudamiento <- filtro_emp_df %>% 
  filter(company != "Resto") %>%
  ggplot(aes(year, liabilities, color = company))+
  geom_point(alpha = 1, size = 3)+
  geom_line()+
  labs(title =  "Liabilities", y = "USD million")+
  theme(legend.position="bottom")
  # scale_y_continuous(labels = scales::percent)+

endeudamiento

# writexl::write_xlsx(list(nocs_data = nocs_df ,
#                          tg_nocs= tg_media), path =  "/Archivos/repos/hidrocarburos/analisis/resultados/argentina/NOC's.xlsx")

#cluster

data_cluster <- nocs_full %>% 
  filter(year == 2011, units == "USD million" ) %>% 
  select(company, productionGroup, indicatorName, observation) %>% 
  mutate(id = 1:nrow(.)) %>% 
  spread(key = indicatorName, value=observation) %>%
  replace(is.na(.), 0 )
  # na.omit()

# data_cluster_z <- scale(data_cluster[4:ncol(data_cluster)])

library(cluster)
datos_para_cluster = data_cluster[4:ncol(data_cluster)]

cantidad_clusters=2

CL  = kmeans(scale(datos_para_cluster),cantidad_clusters)
datos$kmeans = CL$cluster